ARCHIVES

In the modern digital era, organizations increasingly rely on computer networks to store, exchange, and manage sensitive information, which has significantly expanded the landscape of cyber threats. Traditional security mechanisms such as firewalls and signature-based systems are limited in detecting unknown or zero-day attacks. To address these challenges, Intrusion Detection Systems (IDS), particularly Network Intrusion Detection Systems (NIDS), are employed to monitor network traffic and identify malicious activities. In this work, a Network Intrusion Detection System is developed using a hybrid approach that integrates ensemble learning models, namely Random Forest and XGBoost, with a Deep Neural Network (DNN) to enhance detection performance. The system is trained and evaluated using the NSL-KDD dataset, a widely used benchmark in intrusion detection research. Experimental results demonstrate that the proposed system achieves high detection accuracy, with the DNN model attaining approximately 96%, outperforming the other models. The results indicate that combining machine learning and deep learning techniques improves detection accuracy and robustness, making the proposed system suitable for real-world cybersecurity applications.